Trying to follow this thread and wondering if the statement regarding the Metric Push Wheel that measures forward and back to the 1/10 of a meter is correct. If so, wouldn't that be more accurate than the Counter mounted on the bike? I realize the tool is only as good as the user and the process used to measure - but this seems like a simple enough question that from what I read has never been answered.
I have some great pictures of several "certified" steel tapes laid out side by side. There are differences in the readings of +/- 1 cm in places. Just sayin'...
Giles Norton wrote:
I have some great pictures of several "certified" steel tapes laid out side by side. There are differences in the readings of +/- 1 cm in places. Just sayin'...
Giles-
What was the length of the steel tape, surface laid on, and process used to "stretch" each tape. I'm guessing all of the above were equal and your point is that there may be slight errors in the tape measurements/printed increments from the manufacturer?
Another Coach wrote:
Trying to follow this thread and wondering if the statement regarding the Metric Push Wheel that measures forward and back to the 1/10 of a meter is correct. If so, wouldn't that be more accurate than the Counter mounted on the bike?
No.
Giles Norton wrote:
I have some great pictures of several "certified" steel tapes laid out side by side. There are differences in the readings of +/- 1 cm in places. Just sayin'...
http://measure.infopop.cc/eve/forums/a/tpc/f/69510622/m/996306204
Another Coach wrote:
Trying to follow this thread and wondering if the statement regarding the Metric Push Wheel that measures forward and back to the 1/10 of a meter is correct. If so, wouldn't that be more accurate than the Counter mounted on the bike? I realize the tool is only as good as the user and the process used to measure - but this seems like a simple enough question that from what I read has never been answered.
My measuring wheel is direct-drive and shows feet and inches. In a recent test I used it to measure a steel-taped 100-meter section of pavement. It consistently showed 100.5m. Just because it shows you inches doesn't mean it is accurate to within an inch.
scotth wrote:
Another Coach wrote:Trying to follow this thread and wondering if the statement regarding the Metric Push Wheel that measures forward and back to the 1/10 of a meter is correct. If so, wouldn't that be more accurate than the Counter mounted on the bike?
No.
Ditto :)
100.5m CONSISTENTLY proves my point. With the Jones Counter wouldn't you be off several clicks each ride and need to take an average, then recalibrate or check again on the calibration course, then add the SCPF? With that wheel I would think rolling out what you calculate in feet to equal 5025m add 5m for your SCPF and would you really be off by much?
Another Coach wrote:
Trying to follow this thread and wondering if the statement regarding the Metric Push Wheel that measures forward and back to the 1/10 of a meter is correct. If so, wouldn't that be more accurate than the Counter mounted on the bike? I realize the tool is only as good as the user and the process used to measure - but this seems like a simple enough question that from what I read has never been answered.
Devise MW = 10cm
Devise JC = 4.5"
10cm < 4.5"
Devise MW is more accurate
To the OP -
I always wear my Garmin and measure the courses we race on with a wheel. I include in the race summary that I give my kids the wheeled distance. Most courses we compete on are within 20 meters +/- the advertised distance. We do come across a few that are around 100m short. If it is a normal difficulty course and the kids all run large PRs I adjust for the shortness based on their pace/m.
Another Coach wrote:
100.5m CONSISTENTLY proves my point. With the Jones Counter wouldn't you be off several clicks each ride and need to take an average, then recalibrate or check again on the calibration course, then add the SCPF? With that wheel I would think rolling out what you calculate in feet to equal 5025m add 5m for your SCPF and would you really be off by much?
100.5 meters means you were off 1/2 a meter, i.e. 20 inches for each 100 meters.
With a Jones Counter you'd likely get within an inch for each calibration. It's going to come out exactly the same every time, the only difference being if you wobbled around and didn't ride a straight line. The Jones Counter is more accurate than every 4 1/4 inches. That's just how often there are counts. You could mark your wheel around in 80 sections, about 1 inch each, and they would remain the same for the length of the course.
Question:
The paper cited on the previous page had a study on the GPS and issues about where is actually measures. If you measured a course 100 times with the same GPS system and induced no other errors (e.g., stayed on exactly the same line, Would the errors be independent?
Would there be a systematic bias (e.g., in the paper it showed the line going away from the true point, so if you were taking a straight line, the GPS would me measuring a crooked path wavering across the true line)? The net effect would be a slightly longer path. In this case, error along the path, except at the start and finish, would not matter, I do not think. However, those transverse to the path would. If you were 0.5m to the side in a 10m segment, then there is an error of 0.125%. The same error per 20m segment is only 0.031%, and in 50 meters is only 0.005%.
Is this error systematic on a straight line and would be generally systematic (and generally larger on average on a curving course)?
Other errors, if not systematic, can be partially addressed with multiple measurements, it seems. Comments?
Giles Norton wrote:
I have some great pictures of several "certified" steel tapes laid out side by side. There are differences in the readings of +/- 1 cm in places. Just sayin'...
The link shows "normalized" errors. I am guessing that the bigger errors were from shorter tapes, and part of that error can be in the determination of the true starting point, a relative amount the increases the shorter the tape. I bet they were generally not longer than 100m.
In the case of the worst tape: "a length of 30.01 meters" was found for the tape made by/for Stanley in Brazil (good practice would have listed the accuracy to which this measurement is made, such as 30.010 +/-0.001). This might mean that the tape length was between 30.005 and 30.015, so the error then gets 'proportioned' up to 3.1 cms off, but is really in the range of 1.55 4.65 cm off.
What use was made of this tape over time and had it been subject to 'stress' that would stretch it?
I also note that if you take PR's tapes out of the sample, the sample variance plummets, and Pete Riegel is one of the authors, and his co-author is Mike Wickisr, the MW, who has values that are all 99.9x.
This leaves open the question of whether we have observational bias of a type that is common but not always identified. Specifically, the author noticed a flawed tape or some inconsistent tapes. Having made this observation they included the tapes that set off the investigation in the sample. These should have been excluded because they are not a random sample but ones that were 'known' to be aberrant.
Interesting that he conclusion is no problem when the errors could use up a good deal of the built-in caution of 0.01%.
J.R. wrote:
100.5 meters means you were off 1/2 a meter, i.e. 20 inches for each 100 meters.
With a Jones Counter you'd likely get within an inch for each calibration. It's going to come out exactly the same every time, the only difference being if you wobbled around and didn't ride a straight line. The Jones Counter is more accurate than every 4 1/4 inches. That's just how often there are counts. You could mark your wheel around in 80 sections, about 1 inch each, and they would remain the same for the length of the course.
So is there any chance that the .5 off from the push wheel was the actual side to side deviation in the test measurements? Wouldn't a person be able to put marks on the push wheel as well?
Is there really someone out there that can ride a bicycle in a perfectly straigtline for 100m or more?
26mi235 wrote:
Giles Norton wrote:I have some great pictures of several "certified" steel tapes laid out side by side. There are differences in the readings of +/- 1 cm in places. Just sayin'...
The link shows "normalized" errors. I am guessing that the bigger errors were from shorter tapes, and part of that error can be in the determination of the true starting point, a relative amount the increases the shorter the tape. I bet they were generally not longer than 100m.
In the case of the worst tape: "a length of 30.01 meters" was found for the tape made by/for Stanley in Brazil (good practice would have listed the accuracy to which this measurement is made, such as 30.010 +/-0.001). This might mean that the tape length was between 30.005 and 30.015, so the error then gets 'proportioned' up to 3.1 cms off, but is really in the range of 1.55 4.65 cm off.
What use was made of this tape over time and had it been subject to 'stress' that would stretch it?
I own several different brands of steel tapes and some
require different tensions to be considered accurate. Temperature also is a factor to be considered. I whould wager that there are much more noticeable errors in any measuring process than one will find with measuring tool.
Thread Follower wrote:
J.R. wrote:100.5 meters means you were off 1/2 a meter, i.e. 20 inches for each 100 meters.
With a Jones Counter you'd likely get within an inch for each calibration. It's going to come out exactly the same every time, the only difference being if you wobbled around and didn't ride a straight line. The Jones Counter is more accurate than every 4 1/4 inches. That's just how often there are counts. You could mark your wheel around in 80 sections, about 1 inch each, and they would remain the same for the length of the course.
So is there any chance that the .5 off from the push wheel was the actual side to side deviation in the test measurements? Wouldn't a person be able to put marks on the push wheel as well?
Is there really someone out there that can ride a bicycle in a perfectly straigtline for 100m or more?
Good bicyclists can ride a mile on a railroad track; little room for minor deviations and the track is slanted to boot. I could not ride that well, but when training I used to ride on the white line for practice, trying to stay on the line for a mile at a time. Yes, that is following a mark, but it means you get practice riding straight enough for these purposes. Also, if you deviate a couple of cms every 10 meters, the distance difference from a straight line is trivial, since it is a C = (A^2 + B^2)^0.5 and if you are getting a B component in addition to an A component, if the B is small compared to the A component, the difference between A and C is trivial. Riding within 1 part in a 1000 yields an error of 0.00005% and 1 part in 100 is 0.005%, so these are in the trivial range. Also, as for going around corners, you will sometimes be inside and sometimes outside the desired point, so there is a bit of offsetting error there.
Furthermore, the concept you are trying to match is what someone racing could do.
Finally, can I get someone to summarize the rule for XC courses for high school and for college?
I think high school uses the "center of the course" measurement, while college is "1 foot margin around curves"; are these two correct and are there any other factors (other than using the 0.1% margin)?
Course Markings
ARTICLE 3. The course shall be properly measured along the shortest possible route that a runner may take and must be marked clearly by a white or colored line on the ground for the entire route of the course.
Thread Follower wrote:
So is there any chance that the .5 off from the push wheel was the actual side to side deviation in the test measurements? Wouldn't a person be able to put marks on the push wheel as well?
Sure. One issue is that some wheel users seem to think a wheel would not need to be calibrated. If you push a wheel over a calibration 4 times, you should get the same exact measurement each time. That is what the calibration is for. If they wheel wobbles, bounces or bobbles more one time than other, then the calibrations will vary.
Is there really someone out there that can ride a bicycle in a perfectly straigtline for 100m or more?
A bicycle wheel is more likely to keep in a straight line, because of momentum, because of having a constant weight on the wheel, and because the bicycle helps to keep the wheel straight. This is why using the front wheel on a bicycle is more accurate than pushing a wheel.
Sure you could put marks on any wheel, the point being that this does not change the accuracy of the measurement. When the wheel is off 1/2 a meter, it's off 1/2 a meter. When the Jones counter gets the same exact number of counts each time, then the measurement is the same, there is no deviation at all in the measurements.
Look up "The Stability of the Bicycle", by David Jones.
Here is a video that shows an example of David Jones' experiment, note the bigger the trail of a bicycle the greater the stability (and harder to turn). This is why a touring bicycle is better for course measurement than a racing bicycle.
I don't have more exact figures from my own testing as it's been awhile, but the former should have a more normal trail of 2 inches, where a racing might be 3/4 inch or less. This makes a racing bike harder to ride, though some have caught on (like Armstrong) and use a greater trail for greater stability.
Try this with a wheel. :)
http://www.youtube.com/watch?v=PXRQdWG9FuM
The stable/unstable bicycle